Populist Science

• In Event: Expertise in the Age of AI

Abstract

As traditional information-producing organizations, such as news media, government centers, and universities, face unprecedented financial and political threats, how do knowledge professionals determine who and what to trust? This study examines the strategic case of a group of elite academic AI researchers adapting their work under a collapsing peer review ecosystem. Drawing on 11 months of (ongoing) ethnographic observations, 44 interviews, Slack messages, emails, research documents, and other artifacts, I explore how the researchers determine what counts as knowledge from other scientists, and ensure that other scientists in turn consider their work to be knowledge. I find that the researchers adjudicate knowledge almost exclusively through virality on the platform X. Discursively, researchers come to frame virality as a scientific value, justifying the rejection of traditional scientific institutions, such as peer review, in favor of social media fame, which is framed as more meritocratic and inclusive. The counter-intuitive consequence of this virality-based evaluation regime designed for broad participation is that only a small number of people seem to be seen as famous, and thus authoritative, in this political economy of knowledge. I theorize this type of knowledge evaluation practice as populist science. I discuss exceptions and consequences.

Author

• Haley Lepp, Stanford University